Reward function

mapping $\text{gls}fnReward:\text{gls}setActions\times \text{gls}setActions\to \mathbb{R}$ that assigns a scalar value (reward) to each state-action pair; \gls{fnReward} induces the sequence of rewards $\{\text{gls}{rvReward}_t{\}}_{t\in {\mathbb{N}}_0}$, where $\text{gls}{rvReward}_t$ is a \gls{rv} with expected value $\text{gls}fnReward({\mathbf{x}}_t,{\mathbf{a}}_t)$